Sample size and uncertainty
Small samples can swing sharply. A minimum threshold makes a comparison less thin, but it does not establish skill or future profit.
Why a few observations mislead
One win at long odds can dominate ROI in a short series. A few losses can likewise make a record look extremely poor. More settled picks reduce the influence of single events but never remove uncertainty.
HUB requires at least 10 qualifying settled picks for an ROI ranking. Ten is a display threshold, not a statistical seal of quality.
Period and basis of comparison
First ask which period is shown: 14 days, 30 days, 365 days or all time. On the leaderboard, the period is based on kickoff and only picks in the selected window are included.
Prefer comparisons with similar markets, odds and staking profiles. Even a long record may have limited relevance if the approach or data basis has changed.
A cautious checklist
Use the figures to ask better questions, not to predict a certain outcome.
- Read settled pick count alongside ROI and net units.
- Change the period and see whether the conclusion holds.
- Check that both wins and losses are present.
- Never read a minimum sample as a promise of future profit.